Module coa.front
Project : PyCoA - Copyright ©pycoa.fr Date : april 2020 - march 2022 Authors : Olivier Dadoun, Julien Browaeys, Tristan Beau License: See joint LICENSE file
Module : coa.front
About :
This is the PyCoA front end functions. It provides easy access and use of the whole PyCoA framework in a simplified way. The use can change the database, the type of data, the output format with keywords (see help of functions below).
Basic Usage
plotting covid deaths (default value) vs. time import coa.front as cf
cf.plot(where='France') # where keyword is mandatory
getting recovered data for some countries
cf.get(where=['Spain','Italy'],which='recovered')
listing available database and which data can be used cf.listwhom() cf.setwhom('jhu') # return available keywords (aka 'which' data) cf.listwhich() # idem cf.listwhat() # return available time series type (weekly, # daily…) cf.plot(option='sumall') # return the cumulative plot for all countries # for default which keyword. See cf.listwhich() and # and other cf.list**() function (see below)
Expand source code
# -*- coding: utf-8 -*-
"""Project : PyCoA - Copyright ©pycoa.fr
Date : april 2020 - march 2022
Authors : Olivier Dadoun, Julien Browaeys, Tristan Beau
License: See joint LICENSE file
Module : coa.front
About :
-------
This is the PyCoA front end functions. It provides easy access and
use of the whole PyCoA framework in a simplified way.
The use can change the database, the type of data, the output format
with keywords (see help of functions below).
Basic usage
-----------
** plotting covid deaths (default value) vs. time **
import coa.front as cf
cf.plot(where='France') # where keyword is mandatory
** getting recovered data for some countries **
cf.get(where=['Spain','Italy'],which='recovered')
** listing available database and which data can be used **
cf.listwhom()
cf.setwhom('jhu') # return available keywords (aka 'which' data)
cf.listwhich() # idem
cf.listwhat() # return available time series type (weekly,
# daily...)
cf.plot(option='sumall') # return the cumulative plot for all countries
# for default which keyword. See cf.listwhich() and
# and other cf.list**() function (see below)
"""
# --- Imports ----------------------------------------------------------
import pandas as pd
from functools import wraps
import numpy as np
from bokeh.io import show, output_notebook
import types
from coa.tools import kwargs_test, extract_dates, get_db_list_dict, info
import coa.covid19 as coco
from coa.error import *
import coa._version
import coa.geo as coge
output_notebook(hide_banner=True)
# --- Needed global private variables ----------------------------------
_listwhom = list(get_db_list_dict().keys())
if 'coa_db' in __builtins__.keys():
if not __builtins__['coa_db'] in _listwhom:
raise CoaDbError("The variable __builtin__.coa_db set to " + str(__builtins__['coa_db']) +
" which is an invalid db. Error.")
_whom = __builtins__['coa_db']
else:
_whom = _listwhom[0] # default base
_db, _cocoplot = coco.DataBase.factory(_whom) # initialization with default
_gi = None
_dict_bypop = {'no':0,'100':100,'1k':1e3,'100k':1e5,'1M':1e6,'pop':1.}
_listwhat = [ 'standard', # first one is default, nota: we must avoid uppercases
'daily',
'weekly']
_listoutput = ['pandas','geopandas','list', 'dict', 'array'] # first one is default for get
_listvisu = ['bokeh', 'folium']
_listhist = ['bylocation','byvalue','pie']
_listplot = ['date','menulocation','versus','spiral','yearly']
# --- Front end functions ----------------------------------------------
# ----------------------------------------------------------------------
# --- getversion() -----------------------------------------------------
# ----------------------------------------------------------------------
def getversion():
"""Return the current running version of pycoa.
"""
return coa._version.__version__
# ----------------------------------------------------------------------
# --- listoutput() -----------------------------------------------------
# ----------------------------------------------------------------------
def listoutput():
"""Return the list of currently available output types for the
get() function. The first one is the default output given if
not specified.
"""
return _listoutput
# ----------------------------------------------------------------------
# --- listvisu() -------------------------------------------------------
# ----------------------------------------------------------------------
def listvisu():
"""Return the list of currently available visualization for the
map() function. The first one is the default output given if
not specified.
"""
return _listvisu
# ----------------------------------------------------------------------
# --- listwhom() -------------------------------------------------------
# ----------------------------------------------------------------------
def listwhom(detailed=False):
"""Return the list of currently avalailable databases for covid19
data in PyCoA.
The first one is the default one.
If detailed=True, gives information location of each given database.
"""
try:
if int(detailed):
df = pd.DataFrame(get_db_list_dict())
df = df.T.reset_index()
df.index = df.index+1
df = df.rename(columns={'index':'Database',0: "WW/iso3",1:'Granularité',2:'WW/Name'})
return df
else:
return _db.get_available_database()
except:
raise CoaKeyError('Waiting for a boolean !')
# ----------------------------------------------------------------------
# --- listwhat() -------------------------------------------------------
# ----------------------------------------------------------------------
def listwhat():
"""Return the list of currently avalailable type of series available.
The first one is the default one.
"""
return _listwhat
# ----------------------------------------------------------------------
# --- listhist() -------------------------------------------------------
# ----------------------------------------------------------------------
def listhist():
"""Return the list of currently avalailable type of hist available.
The first one is the default one.
"""
return _listhist
# ----------------------------------------------------------------------
# --- listwhat() -------------------------------------------------------
# ----------------------------------------------------------------------
def listplot():
"""Return the list of currently avalailable type of plots available.
The first one is the default one.
"""
return _listplot
# ----------------------------------------------------------------------
# --- listoption() -----------------------------------------------------
# ----------------------------------------------------------------------
def listoption():
"""Return the list of currently avalailable option apply to data.
Default is no option.
"""
return _db.get_available_options()
# ----------------------------------------------------------------------
# --- listtile() -------------------------------------------------------
# ----------------------------------------------------------------------
def listtile():
"""Return the list of currently avalailable tile option for map()
Default is the first one.
"""
return _cocoplot.tiles_list()
# ----------------------------------------------------------------------
# --- listwhich() ------------------------------------------------------
# ----------------------------------------------------------------------
def listwhich():
"""Get which are the available fields for the current base.
Output is a list of string.
By default, the listwhich()[0] is the default which field in other
functions.
"""
return _db.get_available_keys_words()
# ----------------------------------------------------------------------
# --- listwhere() ------------------------------------------------------
# ----------------------------------------------------------------------
def listwhere(clustered = False):
"""Get the list of available regions/subregions managed by the current database
"""
def clust():
if get_db_list_dict()[_whom][1] == 'nation' and get_db_list_dict()[_whom][2] != 'World':
return _db.geo.to_standard(get_db_list_dict()[_whom][0])
else:
r = _db.geo.get_region_list()
if isinstance(r, list):
return r
else:
return sorted(r['name_region'].to_list())
if get_db_list_dict()[_whom][1] == 'nation' and get_db_list_dict()[_whom][2] != 'World':
return [ get_db_list_dict()[_whom][2] ]
if clustered:
return clust()
else:
if _db.db_world == True:
if get_db_list_dict()[_whom][1] == 'nation' and get_db_list_dict()[_whom][2] != 'World':
r = _db.geo.to_standard(get_db_list_dict()[_whom][0])
else:
r = _db.geo.get_GeoRegion().get_countries_from_region('world')
r = [_db.geo.to_standard(c)[0] for c in r]
else:
if get_db_list_dict()[_whom][1] == 'subregion':
pan = _db.geo.get_subregion_list()
r = list(pan.name_subregion.unique())
elif get_db_list_dict()[_whom][1] == 'region':
r = clust()
else:
raise CoaKeyError('What is the granularity of your DB ?')
return r
# ----------------------------------------------------------------------
# --- listbypop() ------------------------------------------------------
# ----------------------------------------------------------------------
def listbypop():
"""Get the list of available population normalization
"""
return list(_dict_bypop.keys())
# ----------------------------------------------------------------------
# --- setwhom() --------------------------------------------------------
# ----------------------------------------------------------------------
def setwhom(base):
"""Set the covid19 database used, given as a string.
Please see pycoa.listbase() for the available current list.
By default, the listbase()[0] is the default base used in other
functions.
"""
global _whom, _db, _cocoplot
if base not in listwhom():
raise CoaDbError(base + ' is not a supported database. '
'See pycoa.listbase() for the full list.')
if True: # force the init #_whom != base:
_db, _cocoplot = coco.DataBase.factory(base)
_whom = base
pass
#return listwhich()
# ----------------------------------------------------------------------
# --- getwhom() --------------------------------------------------------
# ----------------------------------------------------------------------
def getwhom():
"""Return the current base which is used
"""
return _whom
# ----------------------------------------------------------------------
# --- get(**kwargs) ----------------------------------------------------
# ----------------------------------------------------------------------
def getinfo(which):
"""
Return keyword_definition for the db selected
"""
#if not which:
#which = listwhich()[0]
# print('Load default which:',which)
#elif which not in listwhich():
# raise CoaKeyError('Which option ' + which + ' not supported. '
# 'See listwhich() for list.')
print(_db.get_keyword_definition(which),'\nurl:', _db.get_keyword_url(which)[0],'\n(more info ',_db.get_keyword_url(which)[1],')')
# ----------------------------------------------------------------------
# --- Normalisation by pop input pandas return pandas whith by pop new column
# ---------------------------------------------------------------------
def normbypop(pandy, val2norm,bypop):
"""
Return a pandas with a normalisation column add by population
* can normalize by '100', '1k', '100k' or '1M'
"""
global _gi
if pandy.empty:
raise CoaKeyError('Seems to be an empty field')
if isinstance(pandy['codelocation'].iloc[0],list):
pandy = pandy.explode('codelocation')
if _db.db_world == True:
if not isinstance(_gi,coa.geo.GeoInfo):
_gi = coge.GeoInfo()
pop_field='population'
pandy = _gi.add_field(input=pandy,field=pop_field,geofield='codelocation')
else:
if not isinstance(_gi,coa.geo.GeoCountry):
_gi=None
else:
if _gi.get_country() != _db.geo.get_country():
_gi=None
if _gi == None :
_gi = _db.geo
pop_field='population_subregion'
if pop_field not in _gi.get_list_properties():
raise CoaKeyError('The population information not available for this country. No normalization possible')
pandy=_gi.add_field(input=pandy,field=pop_field,input_key='codelocation')
clust = pandy['clustername'].unique()
df = pd.DataFrame()
for i in clust:
pandyi = pandy.loc[ pandy['clustername'] == i ].copy()
pandyi.loc[:,pop_field] = pandyi.groupby('codelocation')[pop_field].first().sum()
if len(pandyi.groupby('codelocation')['codelocation'].first().tolist()) == 1:
cody = pandyi.groupby('codelocation')['codelocation'].first().tolist()*len(pandyi)
else:
cody = [pandyi.groupby('codelocation')['codelocation'].first().tolist()]*len(pandyi)
pandyi = pandyi.assign(codelocation=cody)
if df.empty:
df = pandyi
else:
df = df.append(pandyi)
df = df.drop_duplicates(['date','clustername'])
pandy = df
pandy=pandy.copy()
pandy[pop_field]=pandy[pop_field].replace(0., np.nan)
if bypop == 'pop':
pandy.loc[:,val2norm+' per total population']=pandy[val2norm]/pandy[pop_field]*_dict_bypop[bypop]
else:
pandy.loc[:,val2norm+' per '+bypop + ' population']=pandy[val2norm]/pandy[pop_field]*_dict_bypop[bypop]
return pandy
# ----------------------------------------------------------------------
# --- chartsinput_deco(f)
# ------ with wraps
# ---------- wrapper(*args, **kwargs)
#---------------------------------------
# ----------------------------------------------------------------------
def chartsinput_deco(f):
'''
Main decorator it mainly deals with arg testings
'''
@wraps(f)
def wrapper(*args, **kwargs):
'''
wrapper dealing with arg testing
'''
kwargs_test(kwargs,
['where', 'what', 'which', 'whom', 'when', 'input', 'input_field','output','typeofplot',\
'title','typeofplot','typeofhist','bins','visu','tile','dateslider','maplabel','option','mode','guideline','bypop',
'plot_width','plot_height','textcopyright'],
'Bad args used in the pycoa function.')
# no dateslider currently
global _db, _whom, _gi
where = kwargs.get('where', None)
which = kwargs.get('which', None)
what = kwargs.get('what', _listwhat[0])
whom = kwargs.get('whom', None)
option = kwargs.get('option', None)
when = kwargs.get('when', None)
input_arg = kwargs.get('input', None)
input_field = kwargs.get('input_field',None)
kwargs['input_field'] = None
if which:
kwargs['input_field'] = which
else:
which = listwhich()[0]
if what:
kwargs['input_field'] = what
#if 'mode' in kwargs:
# kwargs.pop('mode')
option = kwargs.get('option', None)
bypop = kwargs.get('bypop','no')
kwargs['output'] = kwargs.get('output', listoutput()[0])
if kwargs['output'] not in listoutput():
raise CoaKeyError('Output option ' + kwargs['output'] + ' not supported. See help().')
if whom is None:
whom = _whom
if whom != _whom:
setwhom(whom)
if not bool([s for s in listwhat() if what.startswith(s)]):
raise CoaKeyError('What option ' + what + ' not supported. '
'See listwhat() for full list.')
#if 'cur_' in which and what == listwhat()[0]:
# what = which
if which not in listwhich():
raise CoaKeyError('Which option ' + which + ' not supported. '
'See listwhich() for list.')
if bypop not in listbypop():
raise CoaKeyError('The bypop arg should be selected in '+str(listbypop)+' only.')
if isinstance(input_arg, pd.DataFrame):
pandy = input_arg #_db.get_stats(which=which, location=where, option=option).rename(columns={'location': 'where'})
if bypop != 'no':
input_arg = normbypop(pandy,input_field,bypop)
if bypop=='pop':
input_field = input_field+' per total population'
else:
input_field = input_field+' per '+ bypop + ' population'
pandy = input_arg
if isinstance(input_field,list):
which = input_field[0]
else:
which = input_field
if which is None:
which = pandy.columns[2]
pandy.loc[:,'standard'] = pandy[which]
kwargs['input_field'] = input_field
if option:
print('option has no effect here, please try do it when you construct your input pandas ')
#input_field = kwargs.get('input_field', listwhich()[0])
#if input_field not in input_arg.columns:
# raise CoaKeyError("Cannot find " + str(input_field) + " field in the pandas data. "
# "Set a proper input_field key.")
elif input_arg == None:
pandy = _db.get_stats(which=which, location=where, option=option).rename(columns={'location': 'where'})
pandy['standard'] = pandy[which]
input_field = what
if bypop != 'no':
if what:
val = what
else:
val = _listwhat[0]
pandy = normbypop(pandy , val, bypop)
if bypop=='pop':
input_field = input_field+' per total population'
else:
input_field = input_field+' per '+ bypop + ' population'
kwargs['input_field'] = input_field
option = kwargs.get('option', None)
else:
raise CoaTypeError('Waiting input as valid pycoa pandas '
'dataframe. See help.')
when_beg, when_end = extract_dates(when)
if pandy[[which,'date']].isnull().values.all():
raise CoaKeyError('All values for '+ which + ' is nan nor empty')
if when_end > pandy[[which,'date']].date.max():
when_end = pandy[[which,'date']].date.max()
db_first_date = pandy[[which,'date']].date.min()
db_last_date = pandy[[which,'date']].date.max()
if when_beg < db_first_date:
when_beg = db_first_date
if when_end > db_last_date:
when_end = db_last_date
# when cut
if when_beg > pandy[[which,'date']].date.max() or when_end > pandy[[which,'date']].date.max():
raise CoaNoData("No available data after "+str( pandy[[which,'date']].date.max()))
pandy = pandy[(pandy.date >= when_beg) & (pandy.date <= when_end)]
if bypop != 'no':
kwargs['input_field'] = [i for i in pandy.columns if ' per ' in i]
#name = [i for i in list(pandy.columns) if ' per ' in i]
if bypop=='pop':
bypop='total population'
else:
bypop+=' population'
if 'tot_' and not what or what=='standard':
renamed = which + ' per '+ bypop
else:
renamed = which + ' '+ what +' per '+ bypop
pandy = pandy.rename(columns={kwargs['input_field'][0]:renamed})
kwargs['input_field'] = renamed
pass
else:
if not input_field or input_field == 'standard':
kwargs['input_field'] = pandy.columns[2]
kwargs['input'] = pandy
return f(**kwargs)
return wrapper
# ----------------------------------------------------------------------
# --- get(**kwargs) ----------------------------------------------------
# ----------------------------------------------------------------------
@chartsinput_deco
def get(**kwargs):
"""Return covid19 data in specified format output (default, by list)
for specified locations ('where' keyword).
The used database is set by the setbase() function but can be
changed on the fly ('whom' keyword)
Keyword arguments
-----------------
where -- a single string of location, or list of (mandatory,
no default value)
which -- what sort of data to deliver ( 'death','confirmed',
'recovered' for 'jhu' default database). See listwhich() function
for full list according to the used database.
what -- which data are computed, either in standard mode
('standard', default value), or 'daily' (diff with previous day
and 'weekly' (diff with previous week). See
listwhich() for fullist of available
Full list of what keyword with the listwhat() function.
whom -- Database specification (overload the setbase()
function). See listwhom() for supported list
when -- dates are given under the format dd/mm/yyyy. In the when
option, one can give one date which will be the end of
the data slice. Or one can give two dates separated with
":", which will define the time cut for the output data
btw those two dates.
output -- output format returned ( pandas (default), array (numpy.array),
dict or list). See listoutput() function.
option -- pre-computing option.
* nonneg means that negative daily balance is pushed back
to previousdays in order to have a cumulative function which is
monotonous increasing.
* nofillnan means that nan value won't be filled.
* smooth7 will perform a 7 day window average of data
* sumall will return integrated over locations given via the
where keyword. If using double bracket notation, the sumall
option is applied for each bracketed member of the where arg.
By default : no option.
See listoption().
bypop -- normalize by population (if available for the selected database).
* by default, 'no' normalization
* can normalize by '100', '1k', '100k' or '1M'
"""
# manage pop norm if asked
#if bypop != 'no':
# if what:
# val2norm=what
# else:
# val2norm=which
# pandy = normbypop(pandy,val2norm,bypop)
# casted_data = None
output = kwargs.get('output')
pandy = kwargs.get('input')
pandy = pandy.drop(columns='standard')
if output == 'pandas':
#if 'option' in kwargs:
# raise CoaKeyError("Cannot use option with input pandas data. "
# "Use option within the get() function instead.")
pandy = pandy.drop_duplicates(['date','clustername'])
#pandy = pandy.drop(columns=['cumul'])
#pandy['cumul'] = pandy[which]
casted_data = pandy
# print(pandy)
# casted_data = pd.pivot_table(pandy, index='date',columns='where',values=col_name).to_dict('series')
# print(pandy)
elif output == 'geopandas':
casted_data = _cocoplot.pycoageo(pandy)
elif output == 'dict':
casted_data = pandy.to_dict('split')
elif output == 'list' or output == 'array':
my_list = []
for keys, values in pandy.items():
vc = [i for i in values]
my_list.append(vc)
casted_data = my_list
if output == 'array':
casted_data = np.array(pandy)
else:
raise CoaKeyError('Unknown output.')
return casted_data
def saveoutput(**kwargs):
'''
Export pycoas pandas as an output file selected by output argument
'pandas': pandas to save
'saveformat': excel (default) or csv
'savename': None (default pycoaout+ '.xlsx/.csv')
'''
global _db
kwargs_test(kwargs, ['pandas','saveformat','savename'], 'Bad args used in the pycoa.saveoutput function.')
pandy = kwargs.get('pandas', pd.DataFrame())
saveformat = kwargs.get('saveformat', 'excel')
savename = kwargs.get('savename', '')
if pandy.empty:
raise CoaKeyError('Pandas to save is mandatory there is not default !')
else:
_db.saveoutput(pandas=pandy,saveformat=saveformat,savename=savename)
def merger(**kwargs):
'''
Merge two or more pycoa pandas from get_stats operation
'coapandas': list (min 2D) of pandas from stats
'whichcol': list variable associate to the coapandas list to be retrieve
'''
global _db
kwargs_test(kwargs,['coapandas'], 'Bad args used in the pycoa.merger function.')
listpandy = kwargs.get('coapandas',[])
return _db.merger(coapandas = listpandy)
# ----------------------------------------------------------------------
# --- map(**kwargs) ----------------------------------------------------
# ----------------------------------------------------------------------
def decomap(func):
def inner(**kwargs):
"""
Create a map according to arguments and options.
See help(map).
- 2 types of visu are avalailable so far : bokeh or folium (see listvisu())
by default visu='bokeh'
- In the default case (i.e visu='bokeh') available option are :
- dateslider=True: a date slider is called and displayed on the right part of the map
- maplabel = text, value are displayed directly on the map
= spark, sparkline are displayed directly on the map
= spiral, spiral are displayed directly on the map
= label%, label are in %
"""
visu = kwargs.get('visu', listvisu()[0])
input = kwargs.get('input')
input_field = kwargs.get('input_field')
if not input.empty:
kwargs.pop('input')
kwargs.pop('input_field')
kwargs.pop('output')
if kwargs.get('bypop'):
kwargs.pop('bypop')
dateslider = kwargs.get('dateslider', None)
maplabel = kwargs.get('maplabel', None)
listmaplabel=['text','spark','spiral','label%','log','unsorted']
if maplabel is not None:
if not isinstance(maplabel,list):
maplabel = [maplabel]
if [a for a in maplabel if a not in listmaplabel]:
raise CoaTypeError('Waiting a correct maplabel value. See help.')
sparkline = False
if dateslider is not None:
del kwargs['dateslider']
kwargs['cursor_date'] = dateslider
if maplabel is not None:
kwargs['maplabel'] = []
if 'text' in maplabel:
kwargs['maplabel'] = ['text']
for i in listmaplabel[1:]:
if i in maplabel:
kwargs['maplabel'].append(i)
#if all([ True if i in ['text','spark','label%','log'] else False for i in kwargs['maplabel'] ]) :
# CoaKeyError('Waiting for a valide label visualisation: text, spark or label%')
input.loc[:,input_field]=input[input_field].fillna(0) #needed in the case where there are nan else js pb
return func(input,input_field,**kwargs)
return inner
@chartsinput_deco
@decomap
def figuremap(input,input_field,**kwargs):
visu = kwargs.get('visu', listvisu()[0])
dateslider = kwargs.get('dateslider', None)
maplabel = kwargs.get('maplabel', None)
if visu == 'bokeh':
if maplabel and 'spark' in maplabel:
return _cocoplot.pycoa_pimpmap(input,input_field,**kwargs)
else:
return _cocoplot.pycoa_map(input,input_field,**kwargs)
@chartsinput_deco
@decomap
def map(input,input_field,**kwargs):
visu = kwargs.get('visu', listvisu()[0])
dateslider = kwargs.get('dateslider', None)
maplabel = kwargs.get('maplabel', None)
if visu == 'bokeh':
if maplabel:
if 'spark' in maplabel or 'spiral' in maplabel:
fig = _cocoplot.pycoa_pimpmap(input,input_field,**kwargs)
elif 'text' in maplabel:
fig = _cocoplot.pycoa_map(input,input_field,**kwargs)
else:
CoaError("What kind of pimp map you want ?!")
else:
fig = _cocoplot.pycoa_map(input,input_field,**kwargs)
return show(fig)
elif visu == 'folium':
if dateslider is not None :
raise CoaKeyError('Not available with folium map, you should considere to use bokeh map visu in this case')
if maplabel and set(maplabel) != set(['log']):
raise CoaKeyError('Not available with folium map, you should considere to use bokeh map visu in this case')
return _cocoplot.pycoa_mapfolium(input,input_field,**kwargs)
else:
raise CoaTypeError('Waiting for a valid visualisation. So far: \'bokeh\' or \'folium\'.See help.')
# ----------------------------------------------------------------------
# --- hist(**kwargs) ---------------------------------------------------
# ----------------------------------------------------------------------
def decohist(func):
def inner(**kwargs):
"""
Create histogram according to arguments.
See help(hist).
Keyword arguments
-----------------
where (mandatory if no input), what, which, whom, when : (see help(get))
input -- input data to plot within the pycoa framework (e.g.
after some analysis or filtering). Default is None which
means that we use the basic raw data through the get
function.
When the 'input' keyword is set, where, what, which,
whom when keywords are ignored.
input should be given as valid pycoa pandas dataframe.
input_field -- is the name of the field of the input pandas to plot.
Default is 'deaths/standard', the default output field of
the get() function.
width_height : width and height of the picture .
If specified should be a list of width and height.
For instance width_height=[400,500]
typeofhist -- 'bylocation' (default), 'byvalue' or pie
bins -- number of bins used, only available for 'byvalue' type of
histograms.
If none provided, a default value will be used.
"""
input = kwargs.pop('input')
input_field = kwargs.pop('input_field')
dateslider = kwargs.get('dateslider', None)
typeofhist = kwargs.pop('typeofhist',listhist()[0])
kwargs.pop('output')
if kwargs.get('bypop'):
kwargs.pop('bypop')
if dateslider is not None:
del kwargs['dateslider']
kwargs['cursor_date'] = dateslider
if typeofhist == 'bylocation':
if 'bins' in kwargs:
raise CoaKeyError("The bins keyword cannot be set with histograms by location. See help.")
fig = _cocoplot.pycoa_horizonhisto(input, input_field, **kwargs)
elif typeofhist == 'byvalue':
if dateslider:
info('dateslider not implemented for typeofhist=\'byvalue\'.')
fig = _cocoplot.pycoa_horizonhisto(input, input_field, **kwargs)
else:
fig = _cocoplot.pycoa_histo(input, input_field, **kwargs)
elif typeofhist == 'pie':
fig = _cocoplot.pycoa_pie(input, input_field, **kwargs)
else:
raise CoaKeyError('Unknown typeofhist value. Available value : listhist().')
return func(fig)
return inner
@chartsinput_deco
@decohist
def figurehist(fig):
''' Return fig Bohek object '''
return fig
@chartsinput_deco
@decohist
def hist(fig):
''' show hist '''
show(fig)
# ----------------------------------------------------------------------
# --- plot(**kwargs) ---------------------------------------------------
# ----------------------------------------------------------------------
def decoplot(func):
def inner(**kwargs):
"""
Create a date plot according to arguments. See help(plot).
Keyword arguments
-----------------
where (mandatory), what, which, whom, when : (see help(get))
input -- input data to plot within the pycoa framework (e.g.
after some analysis or filtering). Default is None which
means that we use the basic raw data through the get
function.
When the 'input' keyword is set, where, what, which,
whom when keywords are ignored.
input should be given as valid pycoa pandas dataframe.
input_field -- is the name of the field of the input pandas to plot.
Default is 'deaths/standard', the default output field of
the get() function.
width_height : width and height of the picture .
If specified should be a list of width and height. For instance width_height=[400,500]
title -- to force the title of the plot
textcopyright - to force the copyright lower left of the graph
typeofplot -- 'date' (default), 'menulocation' or 'versus'
'date':date plot
'spiral': spiral plot if several location only the first one
'menulocation': date plot with two scroll menu locations.
Usefull to study the behaviour of a variable for two different countries.
'versus': plot variable against an other one.
For this type of plot one should used 'input' and 'input_field' (not fully tested).
Moreover dim(input_field) must be 2.
'spiral' : plot variable as a spiral angular plot, angle being the date
'yearly' : same as date but modulo 1 year
"""
input = kwargs.get('input')
input_field = kwargs.get('input_field')
if not input.empty:
kwargs.pop('input')
kwargs.pop('input_field')
typeofplot = kwargs.get('typeofplot',listplot()[0])
kwargs.pop('output')
if kwargs.get('bypop'):
kwargs.pop('bypop')
if 'typeofplot' in kwargs:
typeofplot = kwargs.pop('typeofplot')
if typeofplot == 'date':
fig = _cocoplot.pycoa_date_plot(input,input_field,**kwargs)
elif typeofplot == 'spiral':
fig = _cocoplot.pycoa_spiral_plot(input,input_field,**kwargs)
elif typeofplot == 'versus':
if isinstance(input_field,list) and len(input_field) == 2:
fig = _cocoplot.pycoa_plot(input,input_field,**kwargs)
else:
print('typeofplot is versus but dim(input_field)!=2, versus has not effect ...')
fig = _cocoplot.pycoa_date_plot(input,input_field,**kwargs)
elif typeofplot == 'menulocation':
if get_db_list_dict()[_whom][1] == 'nation' and get_db_list_dict()[_whom][2] != 'World':
print('typeofplot is menulocation with a national DB granularity, use date plot instead ...')
fig = _cocoplot.pycoa_date_plot(input,input_field,**kwargs)
else:
if isinstance(input_field,list) and len(input_field) > 1:
print('typeofplot is menulocation but dim(input_field)>1, menulocation has not effect ...')
fig = _cocoplot.pycoa_scrollingmenu(input,input_field,**kwargs)
elif typeofplot == 'yearly':
fig = _cocoplot.pycoa_yearly_plot(input,input_field,**kwargs)
else:
raise CoaKeyError('Unknown typeofplot value. Should be date, versus, menulocation or spiral.')
return func(fig)
return inner
@chartsinput_deco
@decoplot
def figureplot(fig):
''' Return fig Bohek object '''
return fig
@chartsinput_deco
@decoplot
def plot(fig):
''' show plot '''
show(fig)Functions
def chartsinput_deco(f)-
Main decorator it mainly deals with arg testings
Expand source code
def chartsinput_deco(f): ''' Main decorator it mainly deals with arg testings ''' @wraps(f) def wrapper(*args, **kwargs): ''' wrapper dealing with arg testing ''' kwargs_test(kwargs, ['where', 'what', 'which', 'whom', 'when', 'input', 'input_field','output','typeofplot',\ 'title','typeofplot','typeofhist','bins','visu','tile','dateslider','maplabel','option','mode','guideline','bypop', 'plot_width','plot_height','textcopyright'], 'Bad args used in the pycoa function.') # no dateslider currently global _db, _whom, _gi where = kwargs.get('where', None) which = kwargs.get('which', None) what = kwargs.get('what', _listwhat[0]) whom = kwargs.get('whom', None) option = kwargs.get('option', None) when = kwargs.get('when', None) input_arg = kwargs.get('input', None) input_field = kwargs.get('input_field',None) kwargs['input_field'] = None if which: kwargs['input_field'] = which else: which = listwhich()[0] if what: kwargs['input_field'] = what #if 'mode' in kwargs: # kwargs.pop('mode') option = kwargs.get('option', None) bypop = kwargs.get('bypop','no') kwargs['output'] = kwargs.get('output', listoutput()[0]) if kwargs['output'] not in listoutput(): raise CoaKeyError('Output option ' + kwargs['output'] + ' not supported. See help().') if whom is None: whom = _whom if whom != _whom: setwhom(whom) if not bool([s for s in listwhat() if what.startswith(s)]): raise CoaKeyError('What option ' + what + ' not supported. ' 'See listwhat() for full list.') #if 'cur_' in which and what == listwhat()[0]: # what = which if which not in listwhich(): raise CoaKeyError('Which option ' + which + ' not supported. ' 'See listwhich() for list.') if bypop not in listbypop(): raise CoaKeyError('The bypop arg should be selected in '+str(listbypop)+' only.') if isinstance(input_arg, pd.DataFrame): pandy = input_arg #_db.get_stats(which=which, location=where, option=option).rename(columns={'location': 'where'}) if bypop != 'no': input_arg = normbypop(pandy,input_field,bypop) if bypop=='pop': input_field = input_field+' per total population' else: input_field = input_field+' per '+ bypop + ' population' pandy = input_arg if isinstance(input_field,list): which = input_field[0] else: which = input_field if which is None: which = pandy.columns[2] pandy.loc[:,'standard'] = pandy[which] kwargs['input_field'] = input_field if option: print('option has no effect here, please try do it when you construct your input pandas ') #input_field = kwargs.get('input_field', listwhich()[0]) #if input_field not in input_arg.columns: # raise CoaKeyError("Cannot find " + str(input_field) + " field in the pandas data. " # "Set a proper input_field key.") elif input_arg == None: pandy = _db.get_stats(which=which, location=where, option=option).rename(columns={'location': 'where'}) pandy['standard'] = pandy[which] input_field = what if bypop != 'no': if what: val = what else: val = _listwhat[0] pandy = normbypop(pandy , val, bypop) if bypop=='pop': input_field = input_field+' per total population' else: input_field = input_field+' per '+ bypop + ' population' kwargs['input_field'] = input_field option = kwargs.get('option', None) else: raise CoaTypeError('Waiting input as valid pycoa pandas ' 'dataframe. See help.') when_beg, when_end = extract_dates(when) if pandy[[which,'date']].isnull().values.all(): raise CoaKeyError('All values for '+ which + ' is nan nor empty') if when_end > pandy[[which,'date']].date.max(): when_end = pandy[[which,'date']].date.max() db_first_date = pandy[[which,'date']].date.min() db_last_date = pandy[[which,'date']].date.max() if when_beg < db_first_date: when_beg = db_first_date if when_end > db_last_date: when_end = db_last_date # when cut if when_beg > pandy[[which,'date']].date.max() or when_end > pandy[[which,'date']].date.max(): raise CoaNoData("No available data after "+str( pandy[[which,'date']].date.max())) pandy = pandy[(pandy.date >= when_beg) & (pandy.date <= when_end)] if bypop != 'no': kwargs['input_field'] = [i for i in pandy.columns if ' per ' in i] #name = [i for i in list(pandy.columns) if ' per ' in i] if bypop=='pop': bypop='total population' else: bypop+=' population' if 'tot_' and not what or what=='standard': renamed = which + ' per '+ bypop else: renamed = which + ' '+ what +' per '+ bypop pandy = pandy.rename(columns={kwargs['input_field'][0]:renamed}) kwargs['input_field'] = renamed pass else: if not input_field or input_field == 'standard': kwargs['input_field'] = pandy.columns[2] kwargs['input'] = pandy return f(**kwargs) return wrapper def decohist(func)-
Expand source code
def decohist(func): def inner(**kwargs): """ Create histogram according to arguments. See help(hist). Keyword arguments ----------------- where (mandatory if no input), what, which, whom, when : (see help(get)) input -- input data to plot within the pycoa framework (e.g. after some analysis or filtering). Default is None which means that we use the basic raw data through the get function. When the 'input' keyword is set, where, what, which, whom when keywords are ignored. input should be given as valid pycoa pandas dataframe. input_field -- is the name of the field of the input pandas to plot. Default is 'deaths/standard', the default output field of the get() function. width_height : width and height of the picture . If specified should be a list of width and height. For instance width_height=[400,500] typeofhist -- 'bylocation' (default), 'byvalue' or pie bins -- number of bins used, only available for 'byvalue' type of histograms. If none provided, a default value will be used. """ input = kwargs.pop('input') input_field = kwargs.pop('input_field') dateslider = kwargs.get('dateslider', None) typeofhist = kwargs.pop('typeofhist',listhist()[0]) kwargs.pop('output') if kwargs.get('bypop'): kwargs.pop('bypop') if dateslider is not None: del kwargs['dateslider'] kwargs['cursor_date'] = dateslider if typeofhist == 'bylocation': if 'bins' in kwargs: raise CoaKeyError("The bins keyword cannot be set with histograms by location. See help.") fig = _cocoplot.pycoa_horizonhisto(input, input_field, **kwargs) elif typeofhist == 'byvalue': if dateslider: info('dateslider not implemented for typeofhist=\'byvalue\'.') fig = _cocoplot.pycoa_horizonhisto(input, input_field, **kwargs) else: fig = _cocoplot.pycoa_histo(input, input_field, **kwargs) elif typeofhist == 'pie': fig = _cocoplot.pycoa_pie(input, input_field, **kwargs) else: raise CoaKeyError('Unknown typeofhist value. Available value : listhist().') return func(fig) return inner def decomap(func)-
Expand source code
def decomap(func): def inner(**kwargs): """ Create a map according to arguments and options. See help(map). - 2 types of visu are avalailable so far : bokeh or folium (see listvisu()) by default visu='bokeh' - In the default case (i.e visu='bokeh') available option are : - dateslider=True: a date slider is called and displayed on the right part of the map - maplabel = text, value are displayed directly on the map = spark, sparkline are displayed directly on the map = spiral, spiral are displayed directly on the map = label%, label are in % """ visu = kwargs.get('visu', listvisu()[0]) input = kwargs.get('input') input_field = kwargs.get('input_field') if not input.empty: kwargs.pop('input') kwargs.pop('input_field') kwargs.pop('output') if kwargs.get('bypop'): kwargs.pop('bypop') dateslider = kwargs.get('dateslider', None) maplabel = kwargs.get('maplabel', None) listmaplabel=['text','spark','spiral','label%','log','unsorted'] if maplabel is not None: if not isinstance(maplabel,list): maplabel = [maplabel] if [a for a in maplabel if a not in listmaplabel]: raise CoaTypeError('Waiting a correct maplabel value. See help.') sparkline = False if dateslider is not None: del kwargs['dateslider'] kwargs['cursor_date'] = dateslider if maplabel is not None: kwargs['maplabel'] = [] if 'text' in maplabel: kwargs['maplabel'] = ['text'] for i in listmaplabel[1:]: if i in maplabel: kwargs['maplabel'].append(i) #if all([ True if i in ['text','spark','label%','log'] else False for i in kwargs['maplabel'] ]) : # CoaKeyError('Waiting for a valide label visualisation: text, spark or label%') input.loc[:,input_field]=input[input_field].fillna(0) #needed in the case where there are nan else js pb return func(input,input_field,**kwargs) return inner def decoplot(func)-
Expand source code
def decoplot(func): def inner(**kwargs): """ Create a date plot according to arguments. See help(plot). Keyword arguments ----------------- where (mandatory), what, which, whom, when : (see help(get)) input -- input data to plot within the pycoa framework (e.g. after some analysis or filtering). Default is None which means that we use the basic raw data through the get function. When the 'input' keyword is set, where, what, which, whom when keywords are ignored. input should be given as valid pycoa pandas dataframe. input_field -- is the name of the field of the input pandas to plot. Default is 'deaths/standard', the default output field of the get() function. width_height : width and height of the picture . If specified should be a list of width and height. For instance width_height=[400,500] title -- to force the title of the plot textcopyright - to force the copyright lower left of the graph typeofplot -- 'date' (default), 'menulocation' or 'versus' 'date':date plot 'spiral': spiral plot if several location only the first one 'menulocation': date plot with two scroll menu locations. Usefull to study the behaviour of a variable for two different countries. 'versus': plot variable against an other one. For this type of plot one should used 'input' and 'input_field' (not fully tested). Moreover dim(input_field) must be 2. 'spiral' : plot variable as a spiral angular plot, angle being the date 'yearly' : same as date but modulo 1 year """ input = kwargs.get('input') input_field = kwargs.get('input_field') if not input.empty: kwargs.pop('input') kwargs.pop('input_field') typeofplot = kwargs.get('typeofplot',listplot()[0]) kwargs.pop('output') if kwargs.get('bypop'): kwargs.pop('bypop') if 'typeofplot' in kwargs: typeofplot = kwargs.pop('typeofplot') if typeofplot == 'date': fig = _cocoplot.pycoa_date_plot(input,input_field,**kwargs) elif typeofplot == 'spiral': fig = _cocoplot.pycoa_spiral_plot(input,input_field,**kwargs) elif typeofplot == 'versus': if isinstance(input_field,list) and len(input_field) == 2: fig = _cocoplot.pycoa_plot(input,input_field,**kwargs) else: print('typeofplot is versus but dim(input_field)!=2, versus has not effect ...') fig = _cocoplot.pycoa_date_plot(input,input_field,**kwargs) elif typeofplot == 'menulocation': if get_db_list_dict()[_whom][1] == 'nation' and get_db_list_dict()[_whom][2] != 'World': print('typeofplot is menulocation with a national DB granularity, use date plot instead ...') fig = _cocoplot.pycoa_date_plot(input,input_field,**kwargs) else: if isinstance(input_field,list) and len(input_field) > 1: print('typeofplot is menulocation but dim(input_field)>1, menulocation has not effect ...') fig = _cocoplot.pycoa_scrollingmenu(input,input_field,**kwargs) elif typeofplot == 'yearly': fig = _cocoplot.pycoa_yearly_plot(input,input_field,**kwargs) else: raise CoaKeyError('Unknown typeofplot value. Should be date, versus, menulocation or spiral.') return func(fig) return inner def figurehist(**kwargs)-
Create histogram according to arguments. See help(hist). Keyword Arguments
where (mandatory if no input), what, which, whom, when : (see help(get))
input – input data to plot within the pycoa framework (e.g. after some analysis or filtering). Default is None which means that we use the basic raw data through the get function. When the 'input' keyword is set, where, what, which, whom when keywords are ignored. input should be given as valid pycoa pandas dataframe.
input_field – is the name of the field of the input pandas to plot. Default is 'deaths/standard', the default output field of the get() function.
width_height : width and height of the picture . If specified should be a list of width and height. For instance width_height=[400,500]
typeofhist – 'bylocation' (default), 'byvalue' or pie
bins – number of bins used, only available for 'byvalue' type of histograms. If none provided, a default value will be used.
Expand source code
def inner(**kwargs): """ Create histogram according to arguments. See help(hist). Keyword arguments ----------------- where (mandatory if no input), what, which, whom, when : (see help(get)) input -- input data to plot within the pycoa framework (e.g. after some analysis or filtering). Default is None which means that we use the basic raw data through the get function. When the 'input' keyword is set, where, what, which, whom when keywords are ignored. input should be given as valid pycoa pandas dataframe. input_field -- is the name of the field of the input pandas to plot. Default is 'deaths/standard', the default output field of the get() function. width_height : width and height of the picture . If specified should be a list of width and height. For instance width_height=[400,500] typeofhist -- 'bylocation' (default), 'byvalue' or pie bins -- number of bins used, only available for 'byvalue' type of histograms. If none provided, a default value will be used. """ input = kwargs.pop('input') input_field = kwargs.pop('input_field') dateslider = kwargs.get('dateslider', None) typeofhist = kwargs.pop('typeofhist',listhist()[0]) kwargs.pop('output') if kwargs.get('bypop'): kwargs.pop('bypop') if dateslider is not None: del kwargs['dateslider'] kwargs['cursor_date'] = dateslider if typeofhist == 'bylocation': if 'bins' in kwargs: raise CoaKeyError("The bins keyword cannot be set with histograms by location. See help.") fig = _cocoplot.pycoa_horizonhisto(input, input_field, **kwargs) elif typeofhist == 'byvalue': if dateslider: info('dateslider not implemented for typeofhist=\'byvalue\'.') fig = _cocoplot.pycoa_horizonhisto(input, input_field, **kwargs) else: fig = _cocoplot.pycoa_histo(input, input_field, **kwargs) elif typeofhist == 'pie': fig = _cocoplot.pycoa_pie(input, input_field, **kwargs) else: raise CoaKeyError('Unknown typeofhist value. Available value : listhist().') return func(fig) def figuremap(**kwargs)-
Create a map according to arguments and options. See help(map). - 2 types of visu are avalailable so far : bokeh or folium (see listvisu()) by default visu='bokeh' - In the default case (i.e visu='bokeh') available option are : - dateslider=True: a date slider is called and displayed on the right part of the map - maplabel = text, value are displayed directly on the map = spark, sparkline are displayed directly on the map = spiral, spiral are displayed directly on the map = label%, label are in %
Expand source code
def inner(**kwargs): """ Create a map according to arguments and options. See help(map). - 2 types of visu are avalailable so far : bokeh or folium (see listvisu()) by default visu='bokeh' - In the default case (i.e visu='bokeh') available option are : - dateslider=True: a date slider is called and displayed on the right part of the map - maplabel = text, value are displayed directly on the map = spark, sparkline are displayed directly on the map = spiral, spiral are displayed directly on the map = label%, label are in % """ visu = kwargs.get('visu', listvisu()[0]) input = kwargs.get('input') input_field = kwargs.get('input_field') if not input.empty: kwargs.pop('input') kwargs.pop('input_field') kwargs.pop('output') if kwargs.get('bypop'): kwargs.pop('bypop') dateslider = kwargs.get('dateslider', None) maplabel = kwargs.get('maplabel', None) listmaplabel=['text','spark','spiral','label%','log','unsorted'] if maplabel is not None: if not isinstance(maplabel,list): maplabel = [maplabel] if [a for a in maplabel if a not in listmaplabel]: raise CoaTypeError('Waiting a correct maplabel value. See help.') sparkline = False if dateslider is not None: del kwargs['dateslider'] kwargs['cursor_date'] = dateslider if maplabel is not None: kwargs['maplabel'] = [] if 'text' in maplabel: kwargs['maplabel'] = ['text'] for i in listmaplabel[1:]: if i in maplabel: kwargs['maplabel'].append(i) #if all([ True if i in ['text','spark','label%','log'] else False for i in kwargs['maplabel'] ]) : # CoaKeyError('Waiting for a valide label visualisation: text, spark or label%') input.loc[:,input_field]=input[input_field].fillna(0) #needed in the case where there are nan else js pb return func(input,input_field,**kwargs) def figureplot(**kwargs)-
Create a date plot according to arguments. See help(plot). Keyword Arguments
where (mandatory), what, which, whom, when : (see help(get))
input – input data to plot within the pycoa framework (e.g. after some analysis or filtering). Default is None which means that we use the basic raw data through the get function. When the 'input' keyword is set, where, what, which, whom when keywords are ignored. input should be given as valid pycoa pandas dataframe.
input_field – is the name of the field of the input pandas to plot. Default is 'deaths/standard', the default output field of the get() function.
width_height : width and height of the picture . If specified should be a list of width and height. For instance width_height=[400,500]
title – to force the title of the plot
textcopyright - to force the copyright lower left of the graph
typeofplot – 'date' (default), 'menulocation' or 'versus' 'date':date plot 'spiral': spiral plot if several location only the first one 'menulocation': date plot with two scroll menu locations. Usefull to study the behaviour of a variable for two different countries. 'versus': plot variable against an other one. For this type of plot one should used 'input' and 'input_field' (not fully tested). Moreover dim(input_field) must be 2. 'spiral' : plot variable as a spiral angular plot, angle being the date 'yearly' : same as date but modulo 1 year
Expand source code
def inner(**kwargs): """ Create a date plot according to arguments. See help(plot). Keyword arguments ----------------- where (mandatory), what, which, whom, when : (see help(get)) input -- input data to plot within the pycoa framework (e.g. after some analysis or filtering). Default is None which means that we use the basic raw data through the get function. When the 'input' keyword is set, where, what, which, whom when keywords are ignored. input should be given as valid pycoa pandas dataframe. input_field -- is the name of the field of the input pandas to plot. Default is 'deaths/standard', the default output field of the get() function. width_height : width and height of the picture . If specified should be a list of width and height. For instance width_height=[400,500] title -- to force the title of the plot textcopyright - to force the copyright lower left of the graph typeofplot -- 'date' (default), 'menulocation' or 'versus' 'date':date plot 'spiral': spiral plot if several location only the first one 'menulocation': date plot with two scroll menu locations. Usefull to study the behaviour of a variable for two different countries. 'versus': plot variable against an other one. For this type of plot one should used 'input' and 'input_field' (not fully tested). Moreover dim(input_field) must be 2. 'spiral' : plot variable as a spiral angular plot, angle being the date 'yearly' : same as date but modulo 1 year """ input = kwargs.get('input') input_field = kwargs.get('input_field') if not input.empty: kwargs.pop('input') kwargs.pop('input_field') typeofplot = kwargs.get('typeofplot',listplot()[0]) kwargs.pop('output') if kwargs.get('bypop'): kwargs.pop('bypop') if 'typeofplot' in kwargs: typeofplot = kwargs.pop('typeofplot') if typeofplot == 'date': fig = _cocoplot.pycoa_date_plot(input,input_field,**kwargs) elif typeofplot == 'spiral': fig = _cocoplot.pycoa_spiral_plot(input,input_field,**kwargs) elif typeofplot == 'versus': if isinstance(input_field,list) and len(input_field) == 2: fig = _cocoplot.pycoa_plot(input,input_field,**kwargs) else: print('typeofplot is versus but dim(input_field)!=2, versus has not effect ...') fig = _cocoplot.pycoa_date_plot(input,input_field,**kwargs) elif typeofplot == 'menulocation': if get_db_list_dict()[_whom][1] == 'nation' and get_db_list_dict()[_whom][2] != 'World': print('typeofplot is menulocation with a national DB granularity, use date plot instead ...') fig = _cocoplot.pycoa_date_plot(input,input_field,**kwargs) else: if isinstance(input_field,list) and len(input_field) > 1: print('typeofplot is menulocation but dim(input_field)>1, menulocation has not effect ...') fig = _cocoplot.pycoa_scrollingmenu(input,input_field,**kwargs) elif typeofplot == 'yearly': fig = _cocoplot.pycoa_yearly_plot(input,input_field,**kwargs) else: raise CoaKeyError('Unknown typeofplot value. Should be date, versus, menulocation or spiral.') return func(fig) def get(**kwargs)-
Return covid19 data in specified format output (default, by list) for specified locations ('where' keyword). The used database is set by the setbase() function but can be changed on the fly ('whom' keyword) Keyword Arguments
where – a single string of location, or list of (mandatory, no default value) which – what sort of data to deliver ( 'death','confirmed', 'recovered' for 'jhu' default database). See listwhich() function for full list according to the used database.
what – which data are computed, either in standard mode ('standard', default value), or 'daily' (diff with previous day and 'weekly' (diff with previous week). See listwhich() for fullist of available Full list of what keyword with the listwhat() function.
whom – Database specification (overload the setbase() function). See listwhom() for supported list
when – dates are given under the format dd/mm/yyyy. In the when option, one can give one date which will be the end of the data slice. Or one can give two dates separated with ":", which will define the time cut for the output data btw those two dates.
output – output format returned ( pandas (default), array (numpy.array), dict or list). See listoutput() function.
option – pre-computing option. * nonneg means that negative daily balance is pushed back to previousdays in order to have a cumulative function which is monotonous increasing. * nofillnan means that nan value won't be filled. * smooth7 will perform a 7 day window average of data * sumall will return integrated over locations given via the where keyword. If using double bracket notation, the sumall option is applied for each bracketed member of the where arg.
By default : no option. See listoption().bypop – normalize by population (if available for the selected database). * by default, 'no' normalization * can normalize by '100', '1k', '100k' or '1M'
Expand source code
@chartsinput_deco def get(**kwargs): """Return covid19 data in specified format output (default, by list) for specified locations ('where' keyword). The used database is set by the setbase() function but can be changed on the fly ('whom' keyword) Keyword arguments ----------------- where -- a single string of location, or list of (mandatory, no default value) which -- what sort of data to deliver ( 'death','confirmed', 'recovered' for 'jhu' default database). See listwhich() function for full list according to the used database. what -- which data are computed, either in standard mode ('standard', default value), or 'daily' (diff with previous day and 'weekly' (diff with previous week). See listwhich() for fullist of available Full list of what keyword with the listwhat() function. whom -- Database specification (overload the setbase() function). See listwhom() for supported list when -- dates are given under the format dd/mm/yyyy. In the when option, one can give one date which will be the end of the data slice. Or one can give two dates separated with ":", which will define the time cut for the output data btw those two dates. output -- output format returned ( pandas (default), array (numpy.array), dict or list). See listoutput() function. option -- pre-computing option. * nonneg means that negative daily balance is pushed back to previousdays in order to have a cumulative function which is monotonous increasing. * nofillnan means that nan value won't be filled. * smooth7 will perform a 7 day window average of data * sumall will return integrated over locations given via the where keyword. If using double bracket notation, the sumall option is applied for each bracketed member of the where arg. By default : no option. See listoption(). bypop -- normalize by population (if available for the selected database). * by default, 'no' normalization * can normalize by '100', '1k', '100k' or '1M' """ # manage pop norm if asked #if bypop != 'no': # if what: # val2norm=what # else: # val2norm=which # pandy = normbypop(pandy,val2norm,bypop) # casted_data = None output = kwargs.get('output') pandy = kwargs.get('input') pandy = pandy.drop(columns='standard') if output == 'pandas': #if 'option' in kwargs: # raise CoaKeyError("Cannot use option with input pandas data. " # "Use option within the get() function instead.") pandy = pandy.drop_duplicates(['date','clustername']) #pandy = pandy.drop(columns=['cumul']) #pandy['cumul'] = pandy[which] casted_data = pandy # print(pandy) # casted_data = pd.pivot_table(pandy, index='date',columns='where',values=col_name).to_dict('series') # print(pandy) elif output == 'geopandas': casted_data = _cocoplot.pycoageo(pandy) elif output == 'dict': casted_data = pandy.to_dict('split') elif output == 'list' or output == 'array': my_list = [] for keys, values in pandy.items(): vc = [i for i in values] my_list.append(vc) casted_data = my_list if output == 'array': casted_data = np.array(pandy) else: raise CoaKeyError('Unknown output.') return casted_data def getinfo(which)-
Return keyword_definition for the db selected
Expand source code
def getinfo(which): """ Return keyword_definition for the db selected """ #if not which: #which = listwhich()[0] # print('Load default which:',which) #elif which not in listwhich(): # raise CoaKeyError('Which option ' + which + ' not supported. ' # 'See listwhich() for list.') print(_db.get_keyword_definition(which),'\nurl:', _db.get_keyword_url(which)[0],'\n(more info ',_db.get_keyword_url(which)[1],')') def getversion()-
Return the current running version of pycoa.
Expand source code
def getversion(): """Return the current running version of pycoa. """ return coa._version.__version__ def getwhom()-
Return the current base which is used
Expand source code
def getwhom(): """Return the current base which is used """ return _whom def hist(**kwargs)-
Create histogram according to arguments. See help(hist). Keyword Arguments
where (mandatory if no input), what, which, whom, when : (see help(get))
input – input data to plot within the pycoa framework (e.g. after some analysis or filtering). Default is None which means that we use the basic raw data through the get function. When the 'input' keyword is set, where, what, which, whom when keywords are ignored. input should be given as valid pycoa pandas dataframe.
input_field – is the name of the field of the input pandas to plot. Default is 'deaths/standard', the default output field of the get() function.
width_height : width and height of the picture . If specified should be a list of width and height. For instance width_height=[400,500]
typeofhist – 'bylocation' (default), 'byvalue' or pie
bins – number of bins used, only available for 'byvalue' type of histograms. If none provided, a default value will be used.
Expand source code
def inner(**kwargs): """ Create histogram according to arguments. See help(hist). Keyword arguments ----------------- where (mandatory if no input), what, which, whom, when : (see help(get)) input -- input data to plot within the pycoa framework (e.g. after some analysis or filtering). Default is None which means that we use the basic raw data through the get function. When the 'input' keyword is set, where, what, which, whom when keywords are ignored. input should be given as valid pycoa pandas dataframe. input_field -- is the name of the field of the input pandas to plot. Default is 'deaths/standard', the default output field of the get() function. width_height : width and height of the picture . If specified should be a list of width and height. For instance width_height=[400,500] typeofhist -- 'bylocation' (default), 'byvalue' or pie bins -- number of bins used, only available for 'byvalue' type of histograms. If none provided, a default value will be used. """ input = kwargs.pop('input') input_field = kwargs.pop('input_field') dateslider = kwargs.get('dateslider', None) typeofhist = kwargs.pop('typeofhist',listhist()[0]) kwargs.pop('output') if kwargs.get('bypop'): kwargs.pop('bypop') if dateslider is not None: del kwargs['dateslider'] kwargs['cursor_date'] = dateslider if typeofhist == 'bylocation': if 'bins' in kwargs: raise CoaKeyError("The bins keyword cannot be set with histograms by location. See help.") fig = _cocoplot.pycoa_horizonhisto(input, input_field, **kwargs) elif typeofhist == 'byvalue': if dateslider: info('dateslider not implemented for typeofhist=\'byvalue\'.') fig = _cocoplot.pycoa_horizonhisto(input, input_field, **kwargs) else: fig = _cocoplot.pycoa_histo(input, input_field, **kwargs) elif typeofhist == 'pie': fig = _cocoplot.pycoa_pie(input, input_field, **kwargs) else: raise CoaKeyError('Unknown typeofhist value. Available value : listhist().') return func(fig) def listbypop()-
Get the list of available population normalization
Expand source code
def listbypop(): """Get the list of available population normalization """ return list(_dict_bypop.keys()) def listhist()-
Return the list of currently avalailable type of hist available. The first one is the default one.
Expand source code
def listhist(): """Return the list of currently avalailable type of hist available. The first one is the default one. """ return _listhist def listoption()-
Return the list of currently avalailable option apply to data. Default is no option.
Expand source code
def listoption(): """Return the list of currently avalailable option apply to data. Default is no option. """ return _db.get_available_options() def listoutput()-
Return the list of currently available output types for the get() function. The first one is the default output given if not specified.
Expand source code
def listoutput(): """Return the list of currently available output types for the get() function. The first one is the default output given if not specified. """ return _listoutput def listplot()-
Return the list of currently avalailable type of plots available. The first one is the default one.
Expand source code
def listplot(): """Return the list of currently avalailable type of plots available. The first one is the default one. """ return _listplot def listtile()-
Return the list of currently avalailable tile option for map() Default is the first one.
Expand source code
def listtile(): """Return the list of currently avalailable tile option for map() Default is the first one. """ return _cocoplot.tiles_list() def listvisu()-
Return the list of currently available visualization for the map() function. The first one is the default output given if not specified.
Expand source code
def listvisu(): """Return the list of currently available visualization for the map() function. The first one is the default output given if not specified. """ return _listvisu def listwhat()-
Return the list of currently avalailable type of series available. The first one is the default one.
Expand source code
def listwhat(): """Return the list of currently avalailable type of series available. The first one is the default one. """ return _listwhat def listwhere(clustered=False)-
Get the list of available regions/subregions managed by the current database
Expand source code
def listwhere(clustered = False): """Get the list of available regions/subregions managed by the current database """ def clust(): if get_db_list_dict()[_whom][1] == 'nation' and get_db_list_dict()[_whom][2] != 'World': return _db.geo.to_standard(get_db_list_dict()[_whom][0]) else: r = _db.geo.get_region_list() if isinstance(r, list): return r else: return sorted(r['name_region'].to_list()) if get_db_list_dict()[_whom][1] == 'nation' and get_db_list_dict()[_whom][2] != 'World': return [ get_db_list_dict()[_whom][2] ] if clustered: return clust() else: if _db.db_world == True: if get_db_list_dict()[_whom][1] == 'nation' and get_db_list_dict()[_whom][2] != 'World': r = _db.geo.to_standard(get_db_list_dict()[_whom][0]) else: r = _db.geo.get_GeoRegion().get_countries_from_region('world') r = [_db.geo.to_standard(c)[0] for c in r] else: if get_db_list_dict()[_whom][1] == 'subregion': pan = _db.geo.get_subregion_list() r = list(pan.name_subregion.unique()) elif get_db_list_dict()[_whom][1] == 'region': r = clust() else: raise CoaKeyError('What is the granularity of your DB ?') return r def listwhich()-
Get which are the available fields for the current base. Output is a list of string. By default, the listwhich()[0] is the default which field in other functions.
Expand source code
def listwhich(): """Get which are the available fields for the current base. Output is a list of string. By default, the listwhich()[0] is the default which field in other functions. """ return _db.get_available_keys_words() def listwhom(detailed=False)-
Return the list of currently avalailable databases for covid19 data in PyCoA. The first one is the default one.
If detailed=True, gives information location of each given database.
Expand source code
def listwhom(detailed=False): """Return the list of currently avalailable databases for covid19 data in PyCoA. The first one is the default one. If detailed=True, gives information location of each given database. """ try: if int(detailed): df = pd.DataFrame(get_db_list_dict()) df = df.T.reset_index() df.index = df.index+1 df = df.rename(columns={'index':'Database',0: "WW/iso3",1:'Granularité',2:'WW/Name'}) return df else: return _db.get_available_database() except: raise CoaKeyError('Waiting for a boolean !') def map(**kwargs)-
Create a map according to arguments and options. See help(map). - 2 types of visu are avalailable so far : bokeh or folium (see listvisu()) by default visu='bokeh' - In the default case (i.e visu='bokeh') available option are : - dateslider=True: a date slider is called and displayed on the right part of the map - maplabel = text, value are displayed directly on the map = spark, sparkline are displayed directly on the map = spiral, spiral are displayed directly on the map = label%, label are in %
Expand source code
def inner(**kwargs): """ Create a map according to arguments and options. See help(map). - 2 types of visu are avalailable so far : bokeh or folium (see listvisu()) by default visu='bokeh' - In the default case (i.e visu='bokeh') available option are : - dateslider=True: a date slider is called and displayed on the right part of the map - maplabel = text, value are displayed directly on the map = spark, sparkline are displayed directly on the map = spiral, spiral are displayed directly on the map = label%, label are in % """ visu = kwargs.get('visu', listvisu()[0]) input = kwargs.get('input') input_field = kwargs.get('input_field') if not input.empty: kwargs.pop('input') kwargs.pop('input_field') kwargs.pop('output') if kwargs.get('bypop'): kwargs.pop('bypop') dateslider = kwargs.get('dateslider', None) maplabel = kwargs.get('maplabel', None) listmaplabel=['text','spark','spiral','label%','log','unsorted'] if maplabel is not None: if not isinstance(maplabel,list): maplabel = [maplabel] if [a for a in maplabel if a not in listmaplabel]: raise CoaTypeError('Waiting a correct maplabel value. See help.') sparkline = False if dateslider is not None: del kwargs['dateslider'] kwargs['cursor_date'] = dateslider if maplabel is not None: kwargs['maplabel'] = [] if 'text' in maplabel: kwargs['maplabel'] = ['text'] for i in listmaplabel[1:]: if i in maplabel: kwargs['maplabel'].append(i) #if all([ True if i in ['text','spark','label%','log'] else False for i in kwargs['maplabel'] ]) : # CoaKeyError('Waiting for a valide label visualisation: text, spark or label%') input.loc[:,input_field]=input[input_field].fillna(0) #needed in the case where there are nan else js pb return func(input,input_field,**kwargs) def merger(**kwargs)-
Merge two or more pycoa pandas from get_stats operation 'coapandas': list (min 2D) of pandas from stats 'whichcol': list variable associate to the coapandas list to be retrieve
Expand source code
def merger(**kwargs): ''' Merge two or more pycoa pandas from get_stats operation 'coapandas': list (min 2D) of pandas from stats 'whichcol': list variable associate to the coapandas list to be retrieve ''' global _db kwargs_test(kwargs,['coapandas'], 'Bad args used in the pycoa.merger function.') listpandy = kwargs.get('coapandas',[]) return _db.merger(coapandas = listpandy) def normbypop(pandy, val2norm, bypop)-
Return a pandas with a normalisation column add by population * can normalize by '100', '1k', '100k' or '1M'
Expand source code
def normbypop(pandy, val2norm,bypop): """ Return a pandas with a normalisation column add by population * can normalize by '100', '1k', '100k' or '1M' """ global _gi if pandy.empty: raise CoaKeyError('Seems to be an empty field') if isinstance(pandy['codelocation'].iloc[0],list): pandy = pandy.explode('codelocation') if _db.db_world == True: if not isinstance(_gi,coa.geo.GeoInfo): _gi = coge.GeoInfo() pop_field='population' pandy = _gi.add_field(input=pandy,field=pop_field,geofield='codelocation') else: if not isinstance(_gi,coa.geo.GeoCountry): _gi=None else: if _gi.get_country() != _db.geo.get_country(): _gi=None if _gi == None : _gi = _db.geo pop_field='population_subregion' if pop_field not in _gi.get_list_properties(): raise CoaKeyError('The population information not available for this country. No normalization possible') pandy=_gi.add_field(input=pandy,field=pop_field,input_key='codelocation') clust = pandy['clustername'].unique() df = pd.DataFrame() for i in clust: pandyi = pandy.loc[ pandy['clustername'] == i ].copy() pandyi.loc[:,pop_field] = pandyi.groupby('codelocation')[pop_field].first().sum() if len(pandyi.groupby('codelocation')['codelocation'].first().tolist()) == 1: cody = pandyi.groupby('codelocation')['codelocation'].first().tolist()*len(pandyi) else: cody = [pandyi.groupby('codelocation')['codelocation'].first().tolist()]*len(pandyi) pandyi = pandyi.assign(codelocation=cody) if df.empty: df = pandyi else: df = df.append(pandyi) df = df.drop_duplicates(['date','clustername']) pandy = df pandy=pandy.copy() pandy[pop_field]=pandy[pop_field].replace(0., np.nan) if bypop == 'pop': pandy.loc[:,val2norm+' per total population']=pandy[val2norm]/pandy[pop_field]*_dict_bypop[bypop] else: pandy.loc[:,val2norm+' per '+bypop + ' population']=pandy[val2norm]/pandy[pop_field]*_dict_bypop[bypop] return pandy def plot(**kwargs)-
Create a date plot according to arguments. See help(plot). Keyword Arguments
where (mandatory), what, which, whom, when : (see help(get))
input – input data to plot within the pycoa framework (e.g. after some analysis or filtering). Default is None which means that we use the basic raw data through the get function. When the 'input' keyword is set, where, what, which, whom when keywords are ignored. input should be given as valid pycoa pandas dataframe.
input_field – is the name of the field of the input pandas to plot. Default is 'deaths/standard', the default output field of the get() function.
width_height : width and height of the picture . If specified should be a list of width and height. For instance width_height=[400,500]
title – to force the title of the plot
textcopyright - to force the copyright lower left of the graph
typeofplot – 'date' (default), 'menulocation' or 'versus' 'date':date plot 'spiral': spiral plot if several location only the first one 'menulocation': date plot with two scroll menu locations. Usefull to study the behaviour of a variable for two different countries. 'versus': plot variable against an other one. For this type of plot one should used 'input' and 'input_field' (not fully tested). Moreover dim(input_field) must be 2. 'spiral' : plot variable as a spiral angular plot, angle being the date 'yearly' : same as date but modulo 1 year
Expand source code
def inner(**kwargs): """ Create a date plot according to arguments. See help(plot). Keyword arguments ----------------- where (mandatory), what, which, whom, when : (see help(get)) input -- input data to plot within the pycoa framework (e.g. after some analysis or filtering). Default is None which means that we use the basic raw data through the get function. When the 'input' keyword is set, where, what, which, whom when keywords are ignored. input should be given as valid pycoa pandas dataframe. input_field -- is the name of the field of the input pandas to plot. Default is 'deaths/standard', the default output field of the get() function. width_height : width and height of the picture . If specified should be a list of width and height. For instance width_height=[400,500] title -- to force the title of the plot textcopyright - to force the copyright lower left of the graph typeofplot -- 'date' (default), 'menulocation' or 'versus' 'date':date plot 'spiral': spiral plot if several location only the first one 'menulocation': date plot with two scroll menu locations. Usefull to study the behaviour of a variable for two different countries. 'versus': plot variable against an other one. For this type of plot one should used 'input' and 'input_field' (not fully tested). Moreover dim(input_field) must be 2. 'spiral' : plot variable as a spiral angular plot, angle being the date 'yearly' : same as date but modulo 1 year """ input = kwargs.get('input') input_field = kwargs.get('input_field') if not input.empty: kwargs.pop('input') kwargs.pop('input_field') typeofplot = kwargs.get('typeofplot',listplot()[0]) kwargs.pop('output') if kwargs.get('bypop'): kwargs.pop('bypop') if 'typeofplot' in kwargs: typeofplot = kwargs.pop('typeofplot') if typeofplot == 'date': fig = _cocoplot.pycoa_date_plot(input,input_field,**kwargs) elif typeofplot == 'spiral': fig = _cocoplot.pycoa_spiral_plot(input,input_field,**kwargs) elif typeofplot == 'versus': if isinstance(input_field,list) and len(input_field) == 2: fig = _cocoplot.pycoa_plot(input,input_field,**kwargs) else: print('typeofplot is versus but dim(input_field)!=2, versus has not effect ...') fig = _cocoplot.pycoa_date_plot(input,input_field,**kwargs) elif typeofplot == 'menulocation': if get_db_list_dict()[_whom][1] == 'nation' and get_db_list_dict()[_whom][2] != 'World': print('typeofplot is menulocation with a national DB granularity, use date plot instead ...') fig = _cocoplot.pycoa_date_plot(input,input_field,**kwargs) else: if isinstance(input_field,list) and len(input_field) > 1: print('typeofplot is menulocation but dim(input_field)>1, menulocation has not effect ...') fig = _cocoplot.pycoa_scrollingmenu(input,input_field,**kwargs) elif typeofplot == 'yearly': fig = _cocoplot.pycoa_yearly_plot(input,input_field,**kwargs) else: raise CoaKeyError('Unknown typeofplot value. Should be date, versus, menulocation or spiral.') return func(fig) def saveoutput(**kwargs)-
Export pycoas pandas as an output file selected by output argument 'pandas': pandas to save 'saveformat': excel (default) or csv 'savename': None (default pycoaout+ '.xlsx/.csv')
Expand source code
def saveoutput(**kwargs): ''' Export pycoas pandas as an output file selected by output argument 'pandas': pandas to save 'saveformat': excel (default) or csv 'savename': None (default pycoaout+ '.xlsx/.csv') ''' global _db kwargs_test(kwargs, ['pandas','saveformat','savename'], 'Bad args used in the pycoa.saveoutput function.') pandy = kwargs.get('pandas', pd.DataFrame()) saveformat = kwargs.get('saveformat', 'excel') savename = kwargs.get('savename', '') if pandy.empty: raise CoaKeyError('Pandas to save is mandatory there is not default !') else: _db.saveoutput(pandas=pandy,saveformat=saveformat,savename=savename) def setwhom(base)-
Set the covid19 database used, given as a string. Please see pycoa.listbase() for the available current list.
By default, the listbase()[0] is the default base used in other functions.
Expand source code
def setwhom(base): """Set the covid19 database used, given as a string. Please see pycoa.listbase() for the available current list. By default, the listbase()[0] is the default base used in other functions. """ global _whom, _db, _cocoplot if base not in listwhom(): raise CoaDbError(base + ' is not a supported database. ' 'See pycoa.listbase() for the full list.') if True: # force the init #_whom != base: _db, _cocoplot = coco.DataBase.factory(base) _whom = base pass #return listwhich()